Kernel δ 18 O reflects changes in apical dominance and plant transpiration in tropical maize

Modification of source–sink ratios in tropical maize through detasseling is an ancestral agronomical practice used for increasing yields under stressful conditions. However, the mechanisms behind such effect are not well understood given the difficulties to determine physiological processes such as photosynthesis and whole‐plant transpiration in the field. We have tested the potential ability of kernel δ 18 O to assess differences in grain yield ( GY ) through changes in plant transpiration caused by the modification of water availability and source–sink modification treatments, (including removal of the tassel and different numbers of leaves) in three tropical maize hybrids differing in drought tolerance. Drought‐tolerant genotypes displayed higher yields and lower kernel δ 18 O values than the drought‐susceptible genotype under both well‐watered ( WW ) and water‐stressed ( WS ) conditions. Detasseling caused a positive increase in GY under well‐watered (up to 8%) and water‐deficit conditions (up to 36%). Reduction in leaf area (source) through defoliation treatments caused a large impact on GY showing a trade‐off between maintaining a photosynthetic versus transpiring leaf area. Thus, while a reasonable reduction in leaf area significantly improved plant water availability (as shown by lower kernel δ 18 O values) and consequently GY under water deficit (up to 40%), it caused a maximum reduction of 25% in GY under well‐watered conditions. Variations in GY were significantly ( p < .05) correlated with changes in δ 18 O under both well‐watered ( r = −.67) and WS conditions ( r = .75 and .82). Our results also reinforce the utility of δ 18 O measured in mature kernels as a powerful ecophysiological tool for assessing genotypic differences in apical dominance, transpiration and yield under both WW and WS conditions in tropical maize.